That’s the good stuff. Probably less than 2-5 ppm impurity.
What I read, that’s well above industry standard for “beverage grade!” My problem is I know mine is sourced from a dedicated beverage supply house, but I don’t know if they’re cutting corners. They might figure they’re not supplying a brewery, and a bar that blows a keg in a night could never notice the effects. More investigation ahead.
Buy a $6000 DO meter. Or attack your OCD by taking a risk and trusting them. [emoji16]
Jim, glad to see I can count on you to keep me out of therapy! 
Any time
We’ve thought about trying to capture CO2 blowoff, but it just doesn’t seem feasible. We have a huge tank of liquid CO2 outside that we use for carbonating (and sometimes for moving beer through hoses).
AirGas in Huntsville but I will tell you this: The welding guys said that the big secret is that all the gas comes from the same tank. 
Course, that’s what the welding guys said.
I did a lot of O2 testing on Monday but never got around to testing the welding tanks. I will do that today. I did fix a problem with the keg cleaner(!)
As far as purging tanks goes, I just purge the 5 - 10 psi route 2 or 3 times then purge the head space again and rack beer up from bottom. My DO comes in under 30ppb using this method so not concerned. If I have a really clean fermenting beer like a kolsch or lager I will often add the hops into this tank once the beer has been racked, don’t bother cleaning the tank and racking the dru hopped beers directly to this tank, insuring the least amount of DO pick up.
not sure I follow here?
Good copy
Thanks Keith!
It’s not just welding guys who say that. I have. friend who runs a lab that uses CO2 for very sensitives tests. He’s said the same thing…the only difference is the price you pay.
I’m not following: Everyone get’s higher purity regardless of the price? Or everyone get’s low purity regardless of the price?
Neither of those make sense really. Maybe i’m missing the point somehow.
I will add the hops to the “dirty” tank that is already purged because of racking and run the IPA/hoppy beer into the purged/“dirty” tank. This doesn’t really help a homebrewer unless you have conicals. It’s hard to time that right so only do it with a few beers a year.
now I follow - was reading ‘clean fermenting’ in a different perspective. makes sense now.
My readings varied from 4 ppb to 0 as shown here on the tanks we received directly from the welding company
Wasn’t typing my best this morning, wife up sick coughing all last night. Uggg
But if anyone’s still paranoid I found this:
Interesting!
The amount of oxygen introduced by non-pure CO2 is insignificant to the amount you get by leaving your keg connected to plastic liquid and gas tubing lines. This ingress will happen regardless of how pure your CO2 source is.
You can calculate the amount of atmospheric oxygen that will permeate through the plastic tubing membrane over a given period of time with this equation:
v = P * A * t * (p_1 - p_0) / d
Where
v = the volume of oxygen that diffuses across the membrane
P = the permeability coefficient of the membrane material
A = the area of the membrane
t = time
p_1 = the partial pressure of oxygen in the atmosphere
p_0 = the partial pressure of oxygen inside the tubing
d = the thickness of the membrane
Non-plasticized PVC tubing has a permeability coefficient of P = 0.045 x 10^-10 cm3 (STP)· cm /cm2 · s · cm-Hg (http://www.faybutler.com/pdf_files/HowHoseMaterialsAffectGas3.pdf). Unfortunately, the plasticized version that is required to make flexible tubing has a P that is about 18.75 times higher (http://www.tappi.org/content/events/07place/papers/fischer.pdf). So our P is .84375 x 10^-10 cm3 (STP)· cm /cm2 · s · cm-Hg.
A beverage line with 7/16” = 1.11125 cm OD and 3/16” = 0.47625 cm ID gives you a membrane thickness of 4/16” = 0.635 cm, which is our d. Let’s assume a 1 meter (= 100 cm) tube length which gives an A of pi * OD * 100 = 349.109 cm^2
A gas line with a 9/16” = 1.42875 cm OD and 5/16” ID also has a 4/16” thickness, so we can use the same d = 0.635 cm. The area will be larger: A = pi * 1.42875 * 100 = 448.855 cm^2
Let’s calculate how much oxygen ingress we would get per day. There are 24 * 60 * 60 = 86400 seconds in 1 day, so that’s our t.
p_1 is the partial pressure of oxygen in the atmosphere, which is 0.21 * 76 cm-Hg = 15.96 cm-Hg
p_0 can be assumed to be zero. If it wasn’t there wouldn’t be much point to this exercise in the first place.
Let’s plug it in:
Volume of oxygen that permeates across 1 meter of beer tubing per day:
v = P * A * t * (p_1 - p_0) / d = 0.84375x10^-10 * 349.109 * 86400 * (15.96 - 0) / 0.635 = 0.06396 cm^3 (STP)
cm^3 (STP) is equivalent to the amount of oxygen that would fill one cubic centimeter at standard temperature and pressure. This happens to be 1.43 milligrams worth of oxygen.
Now let’s do the gas line:
v = P * A * t * (p_1 - p_0) / d = 0.84375x10^-10 * 448.855 * 86400 * (15.96 - 0) / 0.635 = 0.08224 cm^3 (STP)
Converting the volumes to weights gives us
0.06396 * 1.43 = 0.09146 milligrams per day through the beer line
0.08224 * 1.43 = 0.11760 milligrams per day through the gas line
But wait. We only considered a 1 meter long length of beer tubing and a 1 meter long length of gas tubing. Most people use a lot more than that! 15 feet seems to be a standard length of beer tubing. Let’s assume 15 feet worth of gas tubing as well. 15 feet is 4.572 meters, so we have to multiply the numbers we found above by 4.572, giving us
0.09146 * 4.572 = 0.41815 milligrams per day through the beer line
0.11760 *4.572 = 0.53767 milligrams per day through the gas line
Alright, so that’s a grand total of 0.95582 milligrams of oxygen per day getting into our beer. Assuming we have 20 liters of beer sitting in the keg, that’s 0.047791 mg/l = 47.79 ppb per day, or 334.5 ppb per week.
There are better materials out there than plasticized PVC, but all of them will leak gas to a certain degree.
I often notice that the same draft beers consistently taste better at certain bars, and worse at certain other bars. I know for a fact that some of them disconnect the tap lines from their kegs overnight.
I buy my co2 from Nexair. They have an online Certificate of Analysis look up tool. I was messing around on with it last night.
You can go here:
http://www.nexair.com/coa/search.php
Enter 393 in the lot number box. The website will return all certificates that start with 393.
Then, lot numbers that end in “211” or “212” are carbon dioxide. The certificates show generally two types of co2 get tested: bone dry and instrument grade. If you look at many certificates you will see that the bone dry purity can vary. Bone dry is usually measured at >99.99% assay but lot 3930016212 was 99.76% assay. The instrument grade always measures >99.99% assay. I recall seeing a pharmaceutical grade certificate at least once. But, I can’t find it again.
The instrument grade certificates measure oxygen content. It is usually listed as <10ppm (LDL)
The Nexair lot number reverse engineered:
ZZZ: 3-digit prefix. Not sure what this means. 392, 393 both return certificates.
WW: looks like a 2-digit week in year. Probably production date because the test date is often a week or two later than that date code.
YY: 2-digit year (16,17,18,…)
XXX: type of gas. CO2 seems to be 211 and 212. My “beverage grade” co2 lot number ends in 211. The certificates always say the product is “Carbon Dioxide - Instrument Grade” or “Carbon Dioxide - Bone Dry”. Both suffixes give both types of products on the certificates.
Edit: my guess is the two types of product listed on the certificates indicate how the product for sale is labeled and that these two labels do not answer the question of whether all the co2 comes from the same source or not.